Exploring the role of a glycine cluster in cold adaptation
of an alkaline phosphatase
Konstantinos Mavromatis
1,
*, Iason Tsigos
2,
*, Maria Tzanodaskalaki
2
, Michael Kokkinidis
1,3
and Vassilis Bouriotis
1,2
1
Department of Biology, Division of Applied Biology and Biotechnology, University of Crete, Greece;
2
Institute of Molecular Biology
and Biotechnology (IMBB), Enzyme Technology Division, and the
3
Institute of Molecular Biology and Biotechnology,
Crystallography Division, Heraklion, Crete, Greece
In an effort to explore the role of glycine clusters on the cold
adaptation of enzymes, we designed point mutations aiming
to alter the distribution of glycine residues close to the active
site of the psychrophilic alkaline phosphatase from the
Antarctic strain TAB5. The mutagenesis targets were
residues Gly261 and Gly262. The replacement of Gly262 by
Ala resulted in an inactive enzyme. Substitution of Gly261
by Ala resulted to an enzyme with lower stability and
increased energy of activation. The double mutant G261A/
Y269A designed on the basis of side-chain packing criteria
from a modelled structure of the enzyme resulted in restor-
ation of the energy of activation to the levels of the native
enzyme and in an increased stability compared to the mutant
G261A. It seems therefore, that the Gly cluster in combi-
nation with its structural environment plays a significant role
in the cold adaptation of the enzyme.
Keywords: alkaline phosphatase; psychrophiles; cold
adaptation; structural flexibility; glycine clusters.
Cold adapted enzymes, produced by organisms living in
permanently cold environments, exhibit a higher specific
activity at low temperatures [1–3]. Moreover, this high
catalytic efficiency is consistently accompanied by a lower
thermal stability, although these properties are not always
correlated as shown by recent data from directed evolution
experiments which support the interdependence of these
properties [4–8].
The adaptation to cold is achieved through a decrease in
the activation energy, which results from an increased
protein flexibility, either of the whole protein or of a specific
domain in some multidomain proteins. Furthermore,
evidence from the notothenioid A4-lactate dehydrogenases
support a cold adaptation model in which structural
flexibility increases are confined to small areas of the
molecule, thereby affecting the mobility of adjacent active
site structures and resulting in reduced energy barriers [9].
Therefore, psychrophilic adaptation seems to be associated
with localized rather than global increases in conformational
flexibility [10]. This is in agreement with structural data,
which reveal that only minor modifications are necessary to
convert a mesophilic or thermophilic enzyme into a cold
adapted one [11–14].
Although the strategy of adaptation is unique to each
enzyme [15], it has been observed that amino-acid residues
involved in the catalytic mechanism are generally conserved
in psychrophilic and mesophilic enzymes [1]. This suggests
that generally the molecular basis of cold adaptation is
associated with sequence changes outside the active site.
However, recent work from our group indicated that the
psychrophilic character of an enzyme could also be altered
or masked by mutating active site residues [16]. Several
sequence patterns have been associated with psychrophilic
adaptations, such as decreased levels of Pro and Arg
residues, weakening of intramolecular interactions,
increased solvent interactions, decreased charged residues
interactions, and disulfide bonds [1,2,17]. Increased levels of
Gly residues or the establishment of Gly clusters have been
frequently suggested to be associated with psychrophilicity
[2]. This could be a result of increased local structural
flexibility due to the intrinsic flexibility of Gly residues [18].
However, recent studies of Gly clusters [19] appear to
contradict this assumption. It seems that the correlation
between the occurrence of Gly residues and the stability of
proteins is complex as several parameters from the whole
protein structure are involved and not just the intrinsic
flexibility of Gly residues [20].
We have recently reported the cloning, sequencing and
overexpression of the gene encoding alkaline phosphatase
from the Antarctic strain TAB5 [16]. Based on the crystal
structure (at 2.4 A
˚)ofanEscherichia coli alkaline phospha-
tase variant with a 28% amino-acid sequence identity to the
psychrophilic enzyme, a three-dimensional model of the
psychrophilic enzyme was constructed [21]. We have also
presented mutagenesis data that substantiate the role of the
local flexibility on the psychrophilic character, and catalytic
properties of the enzyme [16]. In the case of alkaline phos-
phatases, positions 261, 262 (in TAB5 alkaline phosphatase
numbering) are often occupied by one Gly; this site is next
to one of the catalytic residues (Trp260 in the case of TAB5
alkaline phosphatase). In E.coliand some Bacillus sp., there
Correspondence to V. Bouriotis, Department of Biology,
Division of applied Biology and Biotechnology, University of Crete,
PO Box 1470, Heraklion 711 10, Crete, Greece.
Fax/Tel.: + 30 810 394375, E-mail: bouriotis@imbb.forth.gr
Abbreviation:pNPP, p-nitrophenyl phosphate.
Enzyme: alkaline phosphatase (EC 3.1.3.1).
*Note: these authors have equally contributed to this work.
(Received 12 December 2001, revised 14 March 2002,
accepted 18 March 2002)
Eur. J. Biochem. 269, 2330–2335 (2002) FEBS 2002 doi:10.1046/j.1432-1033.2002.02895.x
are no Gly residues at these positions. In TAB5 alkaline
phosphatase, these two positions are both occupied by Gly.
The presence of this Gly cluster in TAB5 alkaline
phosphatase has provoked us to explore its potential role
in the establishment of the psychrophilic properties of the
enzyme.
EXPERIMENTAL PROCEDURES
Materials
Restriction and DNA modification enzymes were pur-
chased from New England Biolabs (Beverly, MA, USA)
and MINOTECH (Heraklion, Greece). All chemicals
were of analytical grade for biochemical use. PCR primers
were purchased from the Microchemistry Laboratory of
IMBB.
Enzymatic assay
Alkaline phosphatase activity was followed spectrophoto-
metrically utilizing p-nitrophenyl phosphate (pNPP) as
substrate. The release of product, p-nitrophenolate, was
monitored by measuring the absorbance at 405 nm using a
PerkinElmer photometer. Specific activity was determined
in a buffer containing 1
M
diethanolamine/HCl (pH 10),
10% glycerol, 10 m
M
MgCl
2
,1m
M
ZnCl
2
,and10m
M
pNPP at 20 C. Enzyme units were calculated as previously
described [22].
Steady-state enzyme kinetics
Steady-state enzyme kinetics were performed in the tem-
perature range 5–25 C. The program
HYPER
v1.01was
used for the determination of V
max
and K
m
values. The k
cat
values were calculated from V
max
using a molecular mass of
76 122 Da for the enzyme. Reported values are the average
of three measurements. The standard deviations do not
exceed 10%. Thermodynamic parameters of the enzyme
were calculated as described previously [27].
Thermal inactivation of enzymes
In order to measure the thermal inactivation of enzymes,
they were incubated at 50 C, in a buffer containing 1
M
diethanolamine pH 10.0, 10 m
M
MgCl
2
,1m
M
ZnCl
2
and
10% glycerol for different time periods and they were
subsequently incubated on ice for 30 min. The remaining
activity was measured at 20 C. Reported values are the
average of at least two measurements. The standard
deviations do not exceed 10%.
Site-directed mutagenesis
Site directed mutagenesis was performed using standard
PCR methods [23]. For the construction of the mutations
the following primers were synthesized: Gly261 to Ala,
upper primer 5¢-d(CAAATAGATTGGGCTGGCCATG
CAAATAAT)-3¢, lower primer 5¢-d(TATTTGCATGGCC
AGCCCAATCTATTTGAG)-3¢; Gly262 to Ala, upper
primer 5¢-d(ATAGATTGGGGTGCCCATGCAAATAA
TGCA)-3¢, lower primer 5¢-d(ATTATTTGCATGGGCA
CCCCAATCTATTTG)-3¢; Tyr269 to Ala, upper primer
5¢-d(TAATGCATCCGCTTTAATTTCTGAAATTA
ATG)-3¢, lower primer 5¢-d(TCAGAAATTAAAGCGG
ATGCATTATTTGCATG)-3¢.
The upstream primer containing the NdeI restriction site
(underlined) was: 5¢-d(GCTAGCATATGAAGCTTAAA
AAAATTG)-3¢and the downstream primer containing the
EcoRI restriction site (underlined) was: 5¢-d(TTGAATTC
GTTTATTGATTCCACTTTG)-3¢.
The PCR reaction mixtures were incubated on an
Eppendorf thermal cycler for 30 cycles of 94 Cfor1min,
49 C for 1 min, and 72 C for 1 min. The amplified
product was isolated by agarose gel electrophoresis, gel
purified using QIAEX (Qiagen) and digested with NdeIand
EcoRI restriction enzymes. The resulting NdeI–EcoRI
fragment was inserted into the pRSETA vector previously
digested with these enzymes. The ligation mixture was used
to transform competent cells of E.colistrain XL1-MRF.
Molecular modelling
A three-dimensional molecular model of the psychrophilic
alkaline phosphatase was built [21] on the basis of the
homology to the E.colienzyme the structure of which is
known [24]. For display of the model and for design and
analysis of mutations the program
SWISSPDB VIEWER
was
used [25].
Expression and purification of enzymes
The protocol used for the expression of enzymes used has
been previously described [16].
RESULTS
Choice of amino-acid substitutions
Based on sequence comparisons, in most alkaline phospha-
tases, the dipeptide corresponding to positions 261 and 262
(TAB5 numbering) contains one Gly residue; the second
residue is usually Ala or His (Fig. 1). In the E.coli
phosphatase these two positions are occupied by Gln and
Asp, respectively. Both positions are occupied by Gly
residues in the TAB5 alkaline phosphatase. This clustering
of Gly provides an interesting mutation target due to its
potential relation to the psychrophilic character of the
enzyme.
Two point mutants were constructed; G261A and
G262A where Gly261 and Gly262 were replaced by Ala,
Fig. 1. Partial alignment of alkaline phosphatases at the region studied.
Mutation targets at positions 261, 262 and 269 of TAB5 alkaline
phosphatase are shown in bold. Grey boxes indicate corresponding
residues in the other alkaline phosphatases.
FEBS 2002 Mutagenesis of a psychrophilic alkaline phosphatase (Eur. J. Biochem. 269) 2331
respectively. By introducing an Ala residue in the place of
Gly it is expected that the conformational flexibility of the
main chain can be constrained with a minimum perturba-
tion of the local structure, resulting to a more rigid protein
(Fig. 2). Moreover, Ala residues are common among
phosphatases at these positions (Fig. 1).
On the basis of the molecular model [21] Ala261 is
expected to introduce steric clashes with the side chain of
Tyr269 (Fig. 2B), which are not present in the structure of
the psychrophilic enzyme with the smaller Gly residue at
position 261 (Fig. 2A). Replacement of Tyr269 by Ala in
the double mutant G261A/Y269A is expected to remove
most of the spatial constraints of the side chain interactions
(Fig. 2C).
Temperature dependence of activity in wild-type
and mutant enzymes
The specific activity of all mutants was measured over the
entire range of temperature (5–25 C) where wild-type
alkaline phosphatase is stable (Fig. 3A). Mutant G262A
had no significant activity at all temperatures tested making
it impossible to measure the specific activity or any kinetic
parameters of this mutant. We could only measure traces of
activity after prolonged incubation (24 h).
The mutant G261A is more active at elevated tempera-
tures (20–25 C) compared to wild-type protein, while the
mutant G261A/Y269A is less active at any given tempera-
ture. However, compared to the mesophilic enzyme from
E.coli, these enzymes are approximately 10 times more
active.
Determination of
E
a
and thermodynamic parameters
for wild-type and mutant enzymes
In order to elucidate the effect of mutations in terms of
psychrophilic adaptation, we determined the energy of
activation E
a
for wild-type and mutant enzymes. Figure 3B
shows the Arrhenius plots for the temperature range of
10–25 C. The E
a
of the enzymes reveal that the mutant
G261A exhibits a higher value almost 2.5-fold higher than
the native cold adapted enzyme (Table 1). The mutant
G261A/Y269A exhibits an E
a
almost the same as in the case
ofthenativeenzyme(Table1).
Thermal inactivation of mutant and wild-type enzymes
In order to investigate the effects of mutations on the
stability of psychrophilic alkaline phosphatase, the enzymes
were incubated at 50 C for different time periods and
subsequently their remaining activity was measured. As
shown in Fig. 3C, replacement of Gly261 by Ala in mutant
G261A resulted in an enzyme with slightly lower stability.
On the other hand, in the double mutant G261A/Y269A the
additional replacement of Tyr269 by Ala restores the
stability of the protein producing a more stable enzyme than
thenativeone.
DISCUSSION
Recent studies have established that, adjustment of con-
formational flexibility is essential for the temperature
adaptation of enzymes [26]. Moreover, localized increases
in conformational flexibility constitute an essential element
in cold adaptation [9]. However, our incomplete under-
standing of the relation between enzyme properties and
conformational flexibility limits the exploitation of the full
potential of protein engineering in the redesign of psychro-
philic enzyme properties [15]. In particular, the effects of
local flexibility in psychrophilic enzyme properties have
been so far studied only for regions, which indirectly affect
the mobility of active site structures, but not for the active
sitesthemselves[9].
Fig. 2. Drawing of the three dimensional model of the wild type (A) and
mutant alkaline phosphatases G261A (B) and G261A/Y269A (C); only
residues that where studied are shown.
2332 K. Mavromatis et al. (Eur. J. Biochem. 269)FEBS 2002
In a previous study [16], we explored the possibility of
modifying the psychrophilic properties of an enzyme by
introducing, via mutagenesis, predictable flexibility changes
to key active site residues of the psychrophilic alkaline
phosphatase from the Antarctic strain TAB5. This
approach was based on an approximate homology-based
three-dimensional model of the psychrophilic enzyme and
sequence comparisons with mesophilic sequences. The
mutagenesis targets were residues Trp260 and Ala219 of
the catalytic site and His135 of the Mg
2+
binding site. The
most striking result was the loss of the psychrophilic
character of mutant W260K/A219N (as reflected by a three-
fold increase of the E
a
value compared to the wild-type
enzyme). Interestingly, the activity of the mutant at elevated
temperatures (20–25 C) exceeded that of the wild-type
protein. Further substitution of His135 by Asp in the triple
mutant W260K/A219N/H135D restored a low energy of
activation. In addition, the His135 fiAsp replacement
resulted in a considerable stabilization of enzymes harboring
this mutation (single mutant H135D and triple mutant
W260K/A219N/H135D). These results suggested that the
psychrophilic character of mutants can be established or
masked by very slight variations of the wild-type sequence,
which may affect various conformational constraints asso-
ciated with active site flexibility.
The aim of the present study was to further explore the
local flexibility concept in the adaptation strategies of
enzymes to low temperatures. As in the previous study [16],
our interest is focused to the vicinity of the active site of the
psychrophilic alkaline phosphatase from the Antarctic
Table 1. Enzymatic and thermodynamic parameters of the psychrophilic alkaline phosphatase and mutants. Reported values were determined at
10 C. The k
cat
values were calculated from V
max
using a molecular weight for the enzyme of 76122 Da in a buffer containing 1
M
diethanolamine-
HCl pH 10, 10% glycerol, 10 m
M
MgCl
2
,1m
M
ZnCl
2
,and10m
M
pNPP. E
a
values were calculated from the slope of the Arrhenius plots in the
temperature range 5–25 C for native and G261A/Y269A mutant, and 5–15 C for the G261A mutant. Thermodynamic parameters DG
#
,DH
#
,
TDS
#
were calculated as described previously [27].
Enzyme
k
cat
(s
)1
)
E
a
(kJÆmol
)1
)
DG
#
(kJÆmol
)1
)
DH
#
(kJÆmol
)1
)
TDS
#
(kJÆmol
)1
)
D(DG
#
)
n-m
(kJÆmol
)1
)
D(DH
#
)
n-m
(kJÆmol
)1
)
TD(DS
#
)
n-m
(kJÆmol
)1
)
Native 1212 42.8 52.48 40.45 )12.03
G261A 423 106.5 54.96 104.15 49.19 )2.48 )63.7 )61.22
G261A/Y269A 310 45.1 55.69 42.75 )12.94 )3.21 )2.3 0.91
Fig. 3. Kinetic studies of wild-type and mutant alkaline phosphatases.
(A) Temperature dependence of k
cat
of TAB5 (d), mutants G261A
(r), G261A/Y269A (j)andE.coli (·) alkaline phosphatases at
temperature range 5–25 C. k
cat
values were determined in a buffer
containing 1
M
diethanolamine-HCl pH 10, 10% glycerol, 10 m
M
MgCl
2
,1m
M
ZnCl
2
,and10m
M
pNPP at 20 C. Alkaline phospha-
tase activity was followed spectrophotometrically utilizing p-nitro-
phenyl phosphate (pNPP) as substrate. The release of product, p-
nitrophenolate, was monitored by measuring the absorbance at
405 nm using a PerkinElmer photometer. Reported values are the
average of three measurements. The standard deviations do not exceed
10%. (B) Arrhenius plots of TAB5, mutants G261A,G261A/Y269A
and E.colialkaline phosphatases. Symbols are as in (A). Reported
values are the average of three measurements. The standard deviations
do not exceed 10%. (C) Thermal inactivation profiles of E.coliand
TAB5 alkaline phosphatases. Enzymes were incubated at 50 C, in a
buffer containing 1
M
diethanolamine pH 10.0, 10 m
M
MgCl
2
,1m
M
ZnCl
2
and 10% glycerol for different time periods and they were
subsequently incubated on ice for 30 min. The remaining activity was
measured at 20 C. Symbols are as in (A). Reported values are the
average of at least two measurements. The standard deviations do not
exceed 10%.
FEBS 2002 Mutagenesis of a psychrophilic alkaline phosphatase (Eur. J. Biochem. 269) 2333
strain TAB5. We particularly attempted to investigate the
functional importance of the Gly pair, located in the vicinity
of the active site of the cold adapted enzyme and to study its
potential role in the establishment of its psychrophilic
character.
This work uses, in accordance with more or less generally
established concepts, the energy of activation, E
a
,asthe
main criterion for the evaluation of the psychrophilic nature
of enzyme variants. In cold adapted enzymes, this param-
eter generally tends to be lower compared to their
mesophilic counterparts [27]. Furthermore, as a measure
of enzyme stability, thermal inactivation at 50 Cisused.
We refer to stability in an activity sense and not in a
thermodynamic sense. We therefore assume that even low
enzymatic activity is associated with a mutant that retains to
a considerable extent the overall fold of the wild-type
protein and that loss of activity is associated either with
perturbation of the native structure or local disruption of
the metal binding or the active site.
The point mutation of Gly262 fiAla results in an
enzyme that exhibits very low activity (less than 1 : 1000
of the native enzyme). This fact did not allow the study
of its kinetic parameters and its thermal inactivation
profile. However, this mutation demonstrates that at
position 262 the presence of Gly is essential, and a
mutation altering this residue results in a practically
inactive enzyme. This Gly may provide the necessary
flexibility required for catalysis. Several alkaline phos-
phatases have one Gly at the corresponding positions
261 and 262, while the psychrophilic enzyme has both
positions occupied by Gly.
The most striking effect of the Gly261 fiAla substitu-
tion (Fig. 2B) is the loss of the psychrophilic character as
deduced from the drastically altered E
a
value (Fig. 3B,
Table 1). As shown in Table 1, this is mainly attributed to
the considerable increase of DH
#
of the mutant compared
to the native enzyme. This observation is in agreement with
previous reports [27], suggesting that the main adaptation of
psychrophilic enzymes lies in a significant decrease of DH
#
with an unavoidable concurrent decrease of TDS
#
.The
slope of the Arrhenius plot, in the temperature range
5–15 C, corresponds to an approximately threefold
increase of the E
a
value compared to the wild-type enzyme.
Interestingly, while this mutant exhibits a considerable
decreased value of k
cat
at lower temperatures, at elevated
temperature (25 C) the value of the same parameter
slightly exceeds that of the wild type (Fig. 3A). This can
be also observed as a bend on the Arrhenius plot occurring
at temperatures > 20 C, indicating that the E
a
value in this
temperature range is considerably lowered. On the basis of
the model, the behavior of the G261A variant can be
interpreted in terms of constraints introduced by the Ala
side chain. The presence of the additional Gly at position
261 possibly offers increased flexibility to the adjacent
residue Trp260 that forms part of the active site and
therefore facilitates the catalysis at low temperatures.
Consequently, when the mutant G261A is driven to operate
in a cold environment, and the lack of Gly261 does not
allow the reaction to proceed as efficiently as in the case of
the native enzyme. At higher temperatures, the additional
energy provided by the environment is sufficient and the
mutant can proceed with the catalysis as efficiently as the
wild type (Fig. 3A). Investigation of the three-dimensional
homology-based model of the enzyme revealed that the
methyl group of Ala261 side-chain could produce steric
clashes with the aromatic ring of Tyr269, and these
unfavorable interactions could lead to a decrease of local
flexibility and an increased E
a
value.
The validity of the above interpretation was further
reinforced by the construction of the double mutant
G261A/Y269A. The additional substitution of Tyr269 fi
Ala was designed with the aim of reducing the spatial
constraints originating from the side-chain interactions
between Tyr269 and Ala261 (Fig. 2C). The main difference
between the G261A and G261A/Y269A enzymes is the
restoration of the psychrophilic character in the double
mutant. Both mutations resulted in an enzyme exhibiting a
significantly lower E
a
,DH
#
and TDS
#
values similar to that
of the wild-type enzyme (Fig. 3B, Table 1). In addition,
considerable stabilization of the double mutant as compared
to the wild-type enzyme was observed (Fig. 3C). This is
probably the result of the relaxationof the side-chain
packing constraints between positions 269 and 261. This
explanation is additionally supported by sequence compar-
isons. As shown in Fig. 1, in other alkaline phosphatases the
corresponding residue at position 269 is often occupied by
residues with smaller side chain when a larger than Gly
residue is found at position 261. This is more striking in the
case of the enzyme from the thermophilic alkaline phos-
phatase from Thermotonga maritima where the presence of
a large side chain (Glu) at corresponding position 261 is
accompanied by a Gly at corresponding position 269 thus
compensating this increase in the side chain volume.
The contribution of Gly clusters in the cold adaptation of
enzymes was also examined in the case of the mammalian
psychrotolerant hormone-sensitive lipase [19]. In that study,
a Gly rich loop (HGGG motif), which was only found in
that enzyme, was extensively mutated and the activity of the
engineered catalysts was analyzed in various temperatures.
However, it was found that although the HGGG loop was a
critical structural element for the catalytic efficiency of the
enzyme, the cold adaptation of the enzyme could not be
attributed to the presence of the Gly cluster in this element.
The present study supports the idea that the Gly cluster,
in combination with its structural environment, is an
essential feature of the psychrophilic character of TAB5
alkaline phosphatase. It seems that the volume of the side-
chains at positions 261 and 269 controls the psychrophilic
character as judged from the levels of the E
a
. In the G261A
mutant, this volume is increased (Fig. 2B) and the enzyme
proves to be as efficient as the native at elevated but not at
lower temperatures. The presence of Gly residues at both
positions 261 and 262 is necessary for the enhanced specific
activity of the enzyme in its natural environment; catalysts
harboring a Gly fiAla mutation in any of these positions
exhibit a significantly decreased specific activity (Fig. 3A).
Consequently, the Gly cluster at this position plays a dual
role, contributing both to higher catalytic efficiency and
lower E
a
.
Moreover, the present work provides evidence that
mutations introduced to Gly cluster produced enzymes that
still exhibit psychrophilic properties while suitable compen-
sating mutations may even produce mutants with increased
stability. To our knowledge, the present study along with a
previous one from our laboratory describing the mutagen-
esis of residues Trp260 and His135 of the same enzyme, are
2334 K. Mavromatis et al. (Eur. J. Biochem. 269)FEBS 2002